Radius based guide selection technique in multi-objective particle swarm optimization

Particle swarm optimization has been shown to solve many real world optimization problems. When PSO is implemented in multi-objective domain, it archives the potential solutions. While solving multi-objective problems using PSO, global best selection from archive impacts on the overall quality of the solution set. In this paper, I present new approach to handle multi-objective problems using PSO. Up till now, the focus of the selection of guide is on the maintaining the spread of the solutions, but in many real-world optimization problems, extreme values are of more interests. I propose the novel to select the global best from archive, which explores the extreme solutions. Performance is compared by applying MOPSO on benchmark functions defined for multi-objective problems.